Shaddows - 11:38 am on Jun 29, 2012 (gmt 0)
claaarky, you have to be careful.
What started as a good bit of empirical investigation is rapidly turning into a splurge of unfounded speculation.
There's an opinion that social marketing affects your rankings. It doesn't, not really.
The cat is out of the bag!
Correlation is not causation. Exit rates are an inverse approximation for user satisfaction (once you normalise for goal completion). Google could be using that raw data. Or, it just might be that they can accuratley predict user disatisfaction.
I mean, users don't leave because the Exit rate is high. They leave because they don't like the page. Google could Pandalise you because you have a high Exit Rate, or because people won't like the page. You can't tell.
What you can do, is use Exit Rate data (again, normalising for goal completion) to establish prime candidates for Pandalisation, and prioritise those for improvement. Or, at the very least, for good honest appraisal.